I’ve watched it happen more times than I’d like to admit. A car rolls in with an intermittent no-start, someone throws a used ECU at it, and three weeks later the same car is back on the same lift with the same fault. The module was never the problem — or rather, the module was the problem, but not in the way anyone assumed. A hairline crack in one solder joint, invisible unless you know exactly what you’re looking for, was quietly opening and closing every time the engine bay got warm.

Solder joint failures are sneaky. They mimic bad grounds, failing relays, corroded terminals, dying crank sensors, and network faults. They come and go with temperature, vibration, and moisture. And they almost never set a helpful fault code.

So here’s the process. Follow it in order. The order matters more than any single test, because each step eliminates a whole category of causes before you go anywhere near a screwdriver.

Step 1: Confirm the Vehicle Complaint (Don’t Skip This)

Before you touch a multimeter, get the story straight. A vague complaint gives you a vague diagnosis. Write down the answers to these:

• Does the vehicle crank? Crank-no-start and no-crank point down completely different roads.
• Does it start intermittently? Intermittent is the classic signature of a mechanical connection failure — and solder joints are mechanical connections.
• Does the failure occur hot or cold? This is gold. Thermal patterns are one of the strongest clues you’ll get.
• Are warning lamps illuminated? Which ones, and do they come on together?
• Can a diagnostic scanner communicate with the ECU? Total comms loss versus normal comms tells you a lot about which part of the board might be involved.
• Does the fault occur after rain or washing? Moisture ingress causes corrosion, and corrosion attacks joints.
• Does vibration affect the fault? Rough roads, speed bumps, engine at idle versus revving — all worth asking about.

A detailed symptom pattern honestly saves hours. I’ve had jobs where the customer’s offhand comment — “it only does it on cold mornings” — pointed straight at the failed joint before the hood was even opened.

Step 2: Check ECU Power and Ground

This is where most misdiagnoses get caught. A poor ground or a corroded power feed can imitate an internal solder failure so convincingly that plenty of good modules get binned over it.

Before opening anything, verify:

• Battery voltage (rested and under crank)
• ECU ignition supply
• Permanent battery supply
• ECU ground circuits
• Main relays
• Relevant fuses
• Voltage drop under load

That last one deserves emphasis. Static voltage readings lie. A ground circuit can read a perfect 0.0 volts of drop with nothing flowing through it and then collapse the moment the starter pulls current. Load-test it. If you’re seeing more than a few hundred millivolts of drop on a ground path under load, fix that before you go any further.

Step 3: Inspect the Wiring and Connectors

Now look at everything between the ECU and the rest of the car:

• Bent terminals
• Loose pins
• Water contamination
• Green or white corrosion
• Damaged wiring
• Pull-out terminals
• Harness tension
• Evidence of overheating (browning, melted insulation, discolored plastic)

Pay special attention to the high-current power and ground terminals. Those carry the most load, generate the most heat, and suffer the most fatigue. If a pin has backed out even slightly, or a terminal has lost its spring tension, you’ll get exactly the same intermittent behavior a cracked joint produces.

Give each wire a gentle tug at the connector body. Not a yank — a firm, controlled pull. A terminal that slides out in your hand just solved your case.

Step 4: Check Communication

If the scanner can’t talk to the ECU, resist the urge to declare the module dead. Test the network first:

• CAN high and CAN low (voltage levels at rest and while active)
• Network resistance (the classic ~60 ohms across a healthy two-terminator bus)
• Short-to-ground faults
• Short-to-power faults
• Other modules on the network — are they all missing, or just this one?
• ECU wake-up and ignition signals

A communication failure can come from the ECU, the wiring, another module dragging the bus down, or a shorted network component. One failed module can silence an entire bus and make five healthy modules look dead. Isolate before you accuse.

Step 5: Perform Controlled Movement Testing

This is the step that finds cracked joints more often than anything else on the list.

With your diagnostic equipment connected and live data or a comms status on screen, carefully apply light pressure to:

• The ECU connector
• The wiring harness
• The ECU housing

Watch for the fault appearing, disappearing, or data glitching. If anything changes, you’ve narrowed it to a connector, terminal, PCB, or solder-joint problem.

Important: do not bend the circuit board or force the connector. Excessive movement creates new damage, and now you’ve got two faults instead of one. Light, deliberate, repeatable pressure. Think “checking a bruise,” not “wrestling a stuck bolt.”

Step 6: Use Temperature Testing — Carefully

A cracked joint is a physics problem. Metal expands when heated and contracts when cooled, and a fractured joint responds to both. Controlled heating or cooling can reproduce an intermittent fault on demand.

What the results tell you:

• Failure when heated: thermal expansion may be opening a cracked joint.
• Failure when cooled: contraction may be separating a weak connection.
• Temporary recovery after heating: a marginal connection expanding back into contact.

Use freeze spray in short bursts and a heat source you can actually control. Avoid uncontrolled heat. Too much temperature will damage processors, memory devices, housing seals, and nearby components — and a heat gun on full blast will happily lift a component off the board and turn a repairable ECU into a paperweight.

Step 7: Open the ECU for Visual Inspection

Only once external causes have been ruled out. Set up properly:

• Magnification (a stereo microscope beats a loupe every time)
• Bright, controlled, angled lighting
• Anti-static protection — wrist strap, mat, grounded bench
• Proper PCB support so the board isn’t flexing under its own weight
• Appropriate cleaning materials (IPA and a soft brush, not household solvents)

Then look for:

• Circular cracks around component leads — the classic signature
• Dull or grainy solder instead of a smooth, shiny fillet
• Ring-shaped fractures
• Lifted pads
• Discolored PCB material
• Burned components
• Corrosion
• Residue (flux, moisture stains, tide marks from water ingress)
• Broken connector solder joints — extremely common, since the connector takes all the mechanical stress
• Cracked large-component joints — relays, capacitors, inductors, power transistors, anything heavy that vibrates

Tilt the board and change your light angle constantly. A crack that’s invisible straight-on will flash into view at 30 degrees.

One honest caveat: a visible crack is excellent evidence, but a joint that looks perfect can still be electrically defective. Cold joints, intermetallic fatigue, and micro-fractures under BGA packages hide beautifully. If the board looks clean but your movement and temperature tests said otherwise, trust the tests.

The Pattern Behind All of This

Every step above does the same job: eliminate the outside world before blaming the inside of the box. Power, grounds, connectors, network, then mechanical stress, then thermal stress, then finally your eyes on the board.

Techs who work in that order fix cars. Techs who skip to Step 7 buy modules.

Where solder joints usually fail is entirely predictable once you’ve seen a few — connector pins, power supply sections, big heavy components, and anywhere the board sees repeated heat cycling. Engine-bay-mounted ECUs suffer worst. Cabin-mounted modules generally live longer, unless water finds them.

Quick Reference Checklist

StepWhat You’re Ruling Out1. Confirm complaintGuesswork and wasted labor2. Power and groundBad feeds imitating internal faults3. Wiring and connectorsCorrosion, backed-out pins, harness damage4. CommunicationBus faults and other modules5. Movement testingMechanical connection failure6. Temperature testingThermal-dependent cracks7. Visual PCB inspectionConfirming the actual joint

Work it top to bottom, document what you find at each stage, and you’ll land on the real fault far more often than the parts-cannon crowd. And when you do finally spot that thin dark ring around a connector pin under the microscope — genuinely one of the more satisfying moments in this trade.